Review and Measurements of Purifi 1ET400A Amplifier

[Julf]

300ohm out and 4.4kΩ in sounds ok?

Pretty OK, depending on how frequency-dependent that 300Ω is.

 

[Armand]

I just remember I saw increased THD when increasing the output impedance from our buffer to the modules. I don't remember how much.

 

[isploit]

https://www.audiosciencereview.com/forum/index.php?pages/Audio_Equipment_Reviews/

I cannot find this in the electronic rankings when I select when I select device type:
-integrated amp
-speaker amp

 

[mk1classic]

 

[txbdan]

Regarding the manufactures using the 1ET400A modules in their designs and offering adjustable gain, I assume that's all in their custom buffer section? Is there a buffer or design that a DIY'er could employ?

 

[isploit]

Regarding the manufactures using the 1ET400A modules in their designs and offering adjustable gain, I assume that's all in their custom buffer section? Is there a buffer or design that a DIY'er could employ?

Eval board comes with purifi buffer. It also has adjustable gain.
VTV seems to offer the cheapest, cheaper as if you were to assemble yourself.

 

[txbdan]

Eval board comes with purifi buffer. It also has adjustable gain.
VTV seems to offer the cheapest, cheaper as if you were to assemble yourself.

Right, but are the manufacturers with a gain knob on the back of their amps using this buffer? The EVAL1 board uses jumpers for discrete gain options or buffer bypass, right? How can a knob be connected? Or maybe they're using a custom buffer with adjustable gain using a pot.

 

[isploit]

Right, but are the manufacturers with a gain knob on the back of their amps using this buffer? The EVAL1 board uses jumpers for discrete gain options or buffer bypass, right? How can a knob be connected? Or maybe they're using a custom buffer with adjustable gain using a pot.

Yes, the eval board uses jumpers for bypassing pre gain. You can see chapter 3.9 of the manual for details:
https://purifi-audio.com/eigentakt/eval1/

I'm cheap so I haven't looked into other manufacturers implementations, I'm sure someone else can answer though..

 

[JimB]

Yes, the eval board uses jumpers for bypassing pre gain. You can see chapter 3.9 of the manual for details:
https://purifi-audio.com/eigentakt/eval1/

I'm cheap so I haven't looked into other manufacturers implementations, I'm sure someone else can answer though..

Okay, you're cheap. So, that should rule out the "adjustable" options - the EVAL1 is the value winner, by far. Do you have a specific different gain in mind? If so, you would change surface mount resistors, as noted here:
https://www.audiosciencereview.com/...iy-purifi-amp-builds.10478/page-8#post-296652
and here:
https://www.audiosciencereview.com/...iy-purifi-amp-builds.10478/page-8#post-296784
Comments about this start from page 2 of this topic.

If you really want a knob, why?

 

[isploit]

Okay, you're cheap. So, that should rule out the "adjustable" options - the EVAL1 is the value winner, by far. Do you have a specific different gain in mind? If so, you would change surface mount resistors, as noted here:
https://www.audiosciencereview.com/...iy-purifi-amp-builds.10478/page-8#post-296652
and here:
https://www.audiosciencereview.com/...iy-purifi-amp-builds.10478/page-8#post-296784
Comments about this start from page 2 of this topic.

If you really want a knob, why?

You got that mixed up^^ @txbdan is inquiring I was just trying to share what little I know about his question.

 

[txbdan]

Okay, you're cheap. So, that should rule out the "adjustable" options - the EVAL1 is the value winner, by far. Do you have a specific different gain in mind? If so, you would change surface mount resistors, as noted here:
https://www.audiosciencereview.com/...iy-purifi-amp-builds.10478/page-8#post-296652
and here:
https://www.audiosciencereview.com/...iy-purifi-amp-builds.10478/page-8#post-296784
Comments about this start from page 2 of this topic.

If you really want a knob, why?

To maximize flexibility and keep my options open in the future. Typical benefits of something with adjustability.

I'm targeting the EVAL1 board so I was wondering if that left my future inflexible in terms of gain. If its still possible to adjust via SMT resistor swaps or even hack in a knob, that's ok with me.

 

[Labjr]

You'd probably need a dual pot for each channel for balanced operation. And pots have tracking issues. Not something I'd do.

 

[JimB]

To maximize flexibility and keep my options open in the future. Typical benefits of something with adjustability.

I see. It is not a particularly common feature. If a need arises for different gain down the road, you could address it then.

 

[JimB]

You'd probably need a dual pot for each channel for balanced operation. And pots have tracking issues. Not something I'd do.

In the cases of the EVAL1 or EVAL2, you don't need a dual per channel. You need to change only the one resistor per channel, so no issue with degrading balance. I'd replace the existing R17 of each channel with the highest value I'd anticipate (lowest gain) and then add resistors off-board to connect in parallel to decrease the resistance and increase gain. The added resistors would be in common on one end of R17 and switched on the other end. I'd use an SPDT switch with a center off position (or an SP3T) right at each stage, and two 1% resistors, each of which would give a different increased gain when paralleled with the value on-board. In this way, you could have no added buffer gain using the jumpers on the board, and three different degrees of added buffer gain with the switched equivalent R17 values. For example, the three switched gain steps could be +6dB (new), +10dB (new) and +14dB (original).

But, like you, not something I am going to do on mine.

 

[txbdan]

I just ordered the VTV EVAL1 "reference" design amp along with a Motu M4 audio interface which outputs up to 5.2Vrms on its balanced outputs. My understanding is that with the default buffer and gain settings, the amp will achieve full output at just under 2Vrms on the balanced inputs. I worry about this discrepancy because it's well over double the max for the amp. I could really goose something if I accidentally run it full blast. Not to mention only using half a knob turn on the M4 makes things a little fiddly.

If I were to disable the EVAL1 "pre" gain stage, my DAC would output about half of what is required to max the amp out. That seems like a better/safer scenario. What are you thoughts and what are the practical cons in disabling the input filtering along with the gain stage? I'm not sure how load sensitive the Motu M4 is, its specs don't mention output impedance.

(This is kinda why a gain knob on the amp would be nice. This won't be my forever DAC.)

I promise I read this entire thread and the DIY thread, but its still a little muddy. Maybe we should make a summary table in the first post.

 

[Julf]

I just ordered the VTV EVAL1 "reference" design amp along with a Motu M4 audio interface which outputs up to 5.2Vrms on its balanced outputs. My understanding is that with the default buffer and gain settings, the amp will achieve full output at just under 2Vrms on the balanced inputs. I worry about this discrepancy because it's well over double the max for the amp. I could really goose something if I accidentally run it full blast. Not to mention only using half a knob turn on the M4 makes things a little fiddly.

That difference is only 6 dB, so you should still have most of the knob turning range available.

 

[isploit]

Would it be a good setup to drive this directly via XLR with Gustard X16? If so would I want to bypass pre-gain?

 

[JimB]

I just ordered the VTV EVAL1 "reference" design amp along with a Motu M4 audio interface which outputs up to 5.2Vrms on its balanced outputs. My understanding is that with the default buffer and gain settings, the amp will achieve full output at just under 2Vrms on the balanced inputs. I worry about this discrepancy because it's well over double the max for the amp. I could really goose something if I accidentally run it full blast. Not to mention only using half a knob turn on the M4 makes things a little fiddly.

If I were to disable the EVAL1 "pre" gain stage, my DAC would output about half of what is required to max the amp out. That seems like a better/safer scenario. What are you thoughts and what are the practical cons in disabling the input filtering along with the gain stage? I'm not sure how load sensitive the Motu M4 is, its specs don't mention output impedance.

(This is kinda why a gain knob on the amp would be nice. This won't be my forever DAC.)

I promise I read this entire thread and the DIY thread, but its still a little muddy. Maybe we should make a summary table in the first post.

I don't understand the anxiety of having an amp sensitivity that only just reaches to clipping at the exact full output of the source. I do understand wanting enough gain to reach rated output (for moments of exuberance), but if the source is capable of a bit more, what would happen? Clipping? If it's too d_mn loud, just turn it down! You might find times you want to turn up some quiet passages and want more gain (within reason). I really can't see the issue with having a source that could reach double the rated input of these amps. Yes, the gain structure gives you the absolute best SN at the point where they are the same, but we have enough here not to need to hit that point, precisely. Am I missing something else?

 

[Deleted member 65]

Just installed/measured/configured Dirac Live Stdalone 3.0 in my Purifi driven system. Had to decrease gain in Dirac Live by -9 dB to avoid clipping (need a new Purifi with double power ;-) ). Gain/Headroom can also be configured in Roon if you’re using that. On my Neurochrome buffers it’s really easy to decrease gain if I want to take that path, just moving a jumper.

 

[txbdan]

I don't understand the anxiety of having an amp sensitivity that only just reaches to clipping at the exact full output of the source. I do understand wanting enough gain to reach rated output (for moments of exuberance), but if the source is capable of a bit more, what would happen? Clipping? If it's too d_mn loud, just turn it down! You might find times you want to turn up some quiet passages and want more gain (within reason). I really can't see the issue with having a source that could reach double the rated input of these amps. Yes, the gain structure gives you the absolute best SN at the point where they are the same, but we have enough here not to need to hit that point, precisely. Am I missing something else?

I'm probably overthinking it. I'll set it up and give it a go in practice. Thanks